Professional Documents
Culture Documents
Megarian Paradoxes As Eleatic Arguments
Megarian Paradoxes As Eleatic Arguments
Megarian Paradoxes As Eleatic Arguments
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
and University of Illinois Press are collaborating with JSTOR to digitize, preserve and extend
access to American Philosophical Quarterly
V. MEGARIAN PARADOXES AS
ELEATIC ARGUMENTS
SAMUEL C. WHEELER HI
287
Megarian solution to the Sorites. Diogenes' attribution supports the claim that
As an historical aside, I will speculate on why Eubulides alone invented all four arguments, but
the Megarians have been so unjustly treated by this is not a central concern of this paper. I will
historians of philosophy. For one thing Diogenes often speak of Eubulides as the author, but I do
Laertius, the main source of information on the this as a convenience.
paradoxes, is a singularly poor person for That even Euclid, let alone Eubulides, was an
transmitting the import of philosophical advocate of Parmenides' ideas at all is not con?
paradoxes. If Russell had been similarly preserved clusively supported by Diogenes' claim that he
for us, we might know only that he thought the read Parmenides' works. Ancient sources besides
present king of France was not bald, that no Diogenes do regard the Megarians as successors of
barbers shaved themselves if they shaved only the Eleatics, but are divided as to what Parmeni?
everybody else and that George IV didn't know dean doctrines they might be committed to.14 The
who the author of Waverly was but did know that main argument I have that the paradoxes are
Scott was Scott. Diogenes seems to have been in? designed to support Parmenidean conclusions is
terested in more "relevant" philosophical views that such a hypothesis gives them a coherent focus
and in words of wisdom rather than in arguments. and makes them into serious problems with
(See his discussion of Plato, for instance.) This metaphysical consequences. If the ascription of
less-than-optimal source partially excuses Parmenidean views to the Megarians turns what
historians such as Copleston, B.A.G. Fuller and the tradition treats as a melange of verbal tricks
"the tradition." into a unified and serious program, that is strong
evidence in favor of the ascription.
II. The Eleaticism of the Megarians
III. Parmenides' Views
The Eleatic point of view here can be made Plato there analyzes negation by the positive
plausible by considering putative negative proper? feature represented by the form Difference.
ties, where properties are thought of as parts of Mourelatos has defended this interpretation at
beings. A negative property is the complement of length and well. In any case, most interpreta?
a positive property. But if properties are real be? tions17 of Parmenides' poem take him to have at
ings, it would be quite remarkable that two pro? least concluded that negation, plurality,
perties should exhaust the world without ever falsehood, and physical existents with parts can?
coinciding. If there were such a complementary not be real phenomena, and these claims are what
property, it could be formally considered a I take the Megarians to be defending.
"negative property," but it would really just be
another positive property ontologically. "Negative IV. The Paradoxes, Their Interpretation, and
properties" are no more real, if properties are the Defense of the Megarian Solutions
parts of the being of things, than negative persons.
("Who came in?" "Well, it was not John." Just as
not-John isn't an individual, so not-red isn't a pro? I will present Eubulides' paradoxes as sets of
perty.) Negative expressions, then, necessarily propositions {A, ...,>U such that each proposi?
lack reference. No part of Being is pictured by any tion A? is intuitively plausible while the set
negative expression. {Au ...,An] is inconsistent. A solution to a
(3) Putative expressions that cannot state how paradox is a rejection of one or more of the pro?
things are cannot be meaningful, since they have positions, say Ai. A solution then says that,
no truth-conditions. By their very nature, they although Ai is very intuitive, ~At is the case. The
can't say "how it is." In particular, ascriptions of theory supported by the paradox is an explanation
"negative properties" cannot picture being, since of why Ai is false. Depending on a theorist's other
there can be no negative properties. commitments, various ^4/s can be given up. That
(4) Since thought is only of what is, there can? is, in general, there is a choice of solutions to a
not be thoughts of putative phenomena that re? paradox which is rationally determined by how
quire negation for their analysis either. strong antecedent grounds for belief in members
(5) Distinctness, plurality, change and motion of [Au ...,An] are. I will argue that there is much
are all putative phenomena whose expression re? to be said for Megarian choices of which A? to
quires negation. So the world cannot have these abandon. At worst, they are defensible as serious
phenomena. The world consists of one, mo? proposals, at best they seem to me absolutely
tionless being. There cannot be falsehood (what is right.
not true), pluralities, or physical objects with As the reader will notice, the explications I pre?
distinct parts (parts which are not one another). sent of the arguments are considerably more
As Parmenides notes,16 it is immaterial which elaborate than the recorded statements of the
propositions are taken to be premises and which arguments. The recorded statements take the form
conclusions. The nonsensicality of the premises of puzzles with no explicitly-drawn conclusions.
and conclusions themselves which is a conse? The hypothesis that these are arguments for
quence of the argument can be tolerated by means Parmenidean conclusions supplies a rationale and
of Wittgenstein's tractarian ladder. (Problems of explanation for their gnomic character.
self-application may explain the "question and Suppose that you wish to argue that there is on?
answer" style of Eubulides, since this avoids com? ly one entity and that concepts which employ
mitments on the part of the interrogator to any negation are inapplicable to reality. If these con?
truth-value for his utterances.) It is worth noting clusions are presented as the conclusions of ex?
in passing that Plato's response to Parmenides in plicitly presented arguments with a plurality of
the Sophist fits Mourelatos' interpretation, since premises using negations, a kind of self-refutation
seems to occur. You would seem to be reaching Zeno takes his paradoxes to show that motion is
impossible, so Eubulides takes this paradox to
the conclusions by nonsense rather than reasons,
if your theory is correct. Your situation would be show that falsehood (as well as "lying" which in?
somewhat like that of the Zen master who must cludes falsehood in its analysis) is an incoherent
teach his students that all dualities are illusory concept. That is, Eubulides takes (l)-(3) to be
without using (and thereby endorsing) dualities. A analytic to the ordinary concepts of falsehood and
Megarian puzzle, I think, is designed to get a per? truth. As Tarski observes, though:23
son to see that some concepts are self "If these observations are correct [i.e., the discus?
contradictory without putting the arguer in the sion of the "universality" of ordinary language],
awkward position of overtly using what must be then the very possibility of a consistent use of the
nonsense. Thus if the puzzles have Eleatic objec? expression "true sentence" which is in harmony with
tives, their form has a point. the laws of logic and the spirit of everyday language
seems to be very questionable ...." Tarski is right
(A) The Liar
only if "not true" is allowed to be true of sentences.
The four groups of versions of the Liar
Eubulides would argue that falsehood cannot be
presented by Bochenski,18 citing Rustow,19 agree
real because, if it were, a self-contradiction would
in presenting the liar as "I am lying," plus a judge?
be the case. Eubulides, on my view, is arguing
ment or question about its truth-value. Familiar
along the lines of Zeno's proof24 that plurality is
presentations20 show how various idiosyncracies
impossible. If a given concept (of a many, of
of this particular version of the paradox can be
falsehood) leads to a contradiction, that concept
eliminated, such as self-reference, token reflex?
must be inapplicable to reality. No proposition is
ives, and so forth, while the paradox remains.
The principles and particular claims below are
expressed by apparent falsehood claims.
Eubulides can rationally give up the intuitively ap?
all intuitively plausible:
parent existence of falsehood because he is con?
(1) Either "true" or "false," but not both, apply
vinced already that there are no "negative proper?
to any statement. A statement is false if and only
if it is not true. ties" and that non-being isn't. In fact, I claim, the
Liar paradox is advanced in order to show that
(2) "I am saying something false" is a mean?
non-being cannot be coherently spoken of, that we
ingful statement, since "Jones is saying something
cannot sensibly say what is not.
false" is meaningful and I could be Jones. "Jones
To recast the argument from Eubulides'
is saying something false" is meaningful, if and
perspective: The concept of falsehood essentially
only if it could say how things are.
involves (l)-(4), but (l)-(4) form an inconsistent
(3) Any language can name any being. So self
set of principles. No concept which is internally in?
reference is not itself pernicious. More generally,
sentences needn't have a determinate meta? consistent can have application to the real world.
Since no paradox is generable if we allow no
linguistic level to be true, so no such requirement
meaningful sentences (i.e., sentences which say
can be put on meaningfulness. For instance, the
how things are) to contain references to non-being
pair of sentences below are both true.
(via negation or concepts such as falsehood), the
The sentence below has six words altogether.
The sentence above contains seven words. sensible alternative is to give up (4), the claim that
there are negative facts about which to speak (i.e.,
(4) Some statements are false. That is, the
the claim that what is not is).
negative property, "not being true" constitutes
part of some state of affairs that B is not true, for (B) The Horned One
some B. This paradox has much the same conclusion and
Kripke21 and others reject (1), Tarski22 rejects motivation as the Liar. In modern terms, this is
(3), and Eubulides rejects (4) and so (2). Just as the problem of accounting for intuitions about
presupposition in a two-valued logic. Historically, we are thinking or speaking sense is not self
this paradox is reported as the sequence (address? evident, at least directly.
ed to a normal person):
(C) The Electra
"What you have not lost, you still have.
"You say Electra knew her brother. But the man
But you have not lost your horns.
at the altar was her brother and she didn't know
So you still have your horns."
him." This argument was familiar to Plato and in
Also, and perhaps more perspicuously:
fact forms one of the central problems of the
"Have you stopped beating your father?
Theatetus. I believe that Plato draws a conclusion
Answer yes or no."
What is paradoxical about these sentences can from the argument similar to ones that Megarians
be set out in the following plausible propositions: draw. I am inclined to ascribe this argument,
(1) If A is a meaningful term and F is a therefore, to Euclid of Megara, since he is a con?
predicate, F(A) is a meaningful statement. temporary of Plato's.
(2) Every meaningful statement S is either true This argument requires more steps and explica?
or not true. When presented as the question, "Is it tion than the previous two. In effect, the first five
the case that S?", every meaningful statement S premises establish that knowledge, if real, is a
has a "yes" or "no" answer. relation. The next three propositions draw conclu?
(3) "Your horns" is a meaningful term, as is sions from this lemma.
"Santa Claus," "What is not," and "Your past (1) The fundamental kind of knowledge is
beating of your father." Generally, meaning knowledge of particulars, that is, cases in which
fulness is self-evident. One can tell when one is "knows" takes a direct object, cases in which the
saying something. state of knowing is directed at an entity. These are
(4) "Have you stopped beating your father?", cases of knowing what individual a thing is.
"Have you lost your horns?" and "Are all of This premise could be supported by any number
John's children blonde?" (when he has none) have of arguments which proceed from the premise that
no "yes" or "no" answers. generalities must be "grounded" on particular
Many logicians deny (4)25, Strawson26 and facts, so that knowledge must do likewise.
others deny (2), and Eubulides denies (3). (2) What a piece of knowledge is knowledge of
Eubulides is committed to denying (3) in any case must be a function of the knower's state. A
in virtue of accepting Parmenides' dictum which knowledge of a particular must be knowledge of
might be translated as "Thought and being are the nothing else, and this must be a feature of the state
same."27 (However this is translated, Parmenides itself.
is at least committed to the view that there are no (3) A knowledge of one truth about an object
thoughts of what isn't. So apparent thoughts, such (or one description under which an object falls)
that, if they are thoughts, are thoughts of what is does not in general suffice for knowledge of that
not, are non-thoughts.) Eubulides, that is, pro? object. The reader, for instance, does not know
duced these problem sentences as devices to con? my maternal grandfather although he knows that
vince us that indeed there cannot be meaningful he fits the description "the author's grandfather."
representation of, i.e., thoughts about, what is (4) Knowledge of every truth about an object
not. (every description under which an object falls)
Eubulides, following Parmenides and like Witt? would suffice to have knowledge of that object.
genstein in the Tractatus, identifies having a sense (5) No intermediate between (3) and (4) seems
(for singular terms) with having a reference. If to be generally formulable in a lawlike way. Cer?
Parmenides is right, and non-being can be neither tain special pieces of information can be found
thought nor said, then many apparent thoughts which, together with other conditions, lead us in?
are not thoughts at all, but rather nonsense. That tuitively to say that a person in those conditions
with these pieces of information knows an entity taken to be an inconsistency (a claim that a rela?
of a given kind, but each kind of case seems dif? tion both holds and fails to hold) rather than a
ferent. (Boer and Lycan28, for instance, spend demonstration that knowledge isn't a relation at
close to fifty pages on "knowing who," which all.
covers just persons.) A full set of general (9) (a) Sense objects, about which there must
necessary and sufficient conditions for knowing be not-now-known truths, can be known. (This is
what a particular is, even if possible, would be a plausible version of what must be dropped as a
enormously complex. It could even be the case result of the Theatetus* similar argument.)
that no finite characterization of "knows what X (b) More generally, objects which bear in?
is" is possible which allows extension to new kinds definite multiplicities of contingent relation to
of case. other objects can be known.
(6) Knowledge of particulars, if a real Euclid and Plato, I claim, give up (9). Thus, on
my account, Euclid advanced the Electra to show
phenomenon, has a simple law-like account. This
that the Parmenidean One is the only possible ob?
is really a point about the philosophy of science. If
ject of knowledge. There are only a finite number
what exists is a cosmos, then all real phenomena
of truths about the one, so that it can be totally
fit together in one whole. A central phenomenon
known, and so known at all. According to Plato,
such as knowledge (in Plato's scheme) must have
the Forms constitute a determinate graspable
intelligible connections with other kinds of whole and do not have relations to contingent
phenomena. The existence of such connecting non-eternal entities.
laws is the mark of being a central part of a Modern philosophers, of course, deny that
cosmos. If electromagnetic fields had no simpler knowledge is a relation by abandoning some of
laws than the account of knowledge which would (l)-(6). For a scientific realist this is tantamount to
fit the varying intuitions we have, it would be irra? denying that knowledge is a real phenomenon. For
tional to postulate their existence. The Greeks, or a scientific realist, all real phenomena are on a
at least Plato and Euclid, saw no reason to treat par. Thus knowledge, as an alleged natural kind
knowledge, if a real phenomenon, any differently of phenomenon, must meet the same kind of stan?
from other phenomena. If there is such a thing, it dards as any other. If intuitions about knowing
has a simple account. We should bear in mind who a person is, knowing what a thing is, and be?
that, for Plato, intentional/mental phenomena ing acquainted with a thing cannot be captured by
are built into the structure of the universe. A fun? some systematization that makes it plausible that
damental psychological concept must also be a knowledge is a natural phenomenon, then those
fundamental natural concept. intuitions don't fit any natural phenomenon.
(7) From (l)-(6) it is plausible that only what is There is then the choice between deciding that
totally known, in the sense that every description some other natural phenomenon is knowledge but
under which it falls is known, can be known at all. that our intuitions have nothing to do with it,
Thus knowledge is a relation of knower and deciding that knowledge is a supernatural
known. So nothing is both known and not known. phenomenon, or deciding that knowledge is not a
(8) Thus in particular, the intuitive belief that real psychological phenomenon but rather that the
Electra knew her brother (since she grew up with concept is a rather inept categorization of human
him) is shown to be inconsistent with the admitted states in relation to the world. Thus I think that
fact that she did not know him at the altar. The Euclid's argument shows that // knowledge is a
"electra" was presented in this form in order to real phenomenon, it is a relation. So either scien?
convince the interlocuter that his intuitive applica? tific realism or the reality of ordinary empirical
tion of "knows" is inconsistent. What premises knowledge must go. Similar arguments, of course,
(l)-(6) do is to give reasons why this should be apply to other psychological attitudes.
The resemblance to Zeno's arguments are clearest is not only false that nature is partitioned into tall
in this state, and in fact his use of divisibility con? things and persons, but impossible.
siderations may have inspired it.
I have applied a somewhat modified form of the
V. Conclusion
sorites argument which drops the presumptions
that inconsistent concepts cannot be true of en?
tities. The "analytic content" of, say, an electron If these interpretations are even roughly correct,
concept may be such that small dimunitions in the Megarians are a vastly under-rated group of
charge leave an electron in existence even though philosophers. I have argued that the arguments at?
this isn't true of electrons. Thus, for me if there tributed to Eubulides are not silly puzzles but
are sharp breaks in the nature of things, a term rather powerful, serious, and plausible defenses of
can apply to things even though its concept doesn't Parmenidean conclusions. Their strange conclu?
"fit" them. Peter Unger, in several recent articles sions detract from their interest no more than
has argued a very pure Eubulideah line in which it those of Zeno's arguments do.
NOTES
1. Frederick Copleston, A History of Philosophy, Vol. 1, Pt. 2, (Doubleday&Co., Inc.: Garden City, New York, 1962), p. 137.
2. Eduard Zeller, in Outline of the History of Greek Philosophy (Harcourt, Brace & Co.: New York, 1931), p. 107.
3. Wilhelm, Windelband, A History of Philosophy, Vol. I (Harper & Brothers: New York, 1958), p. 89.
4. B. A. G. Fuller, History of Greek Philosophy, Vol. II, (Henry Holt and Company: New York, 1931), p. 94.
5. Another writer who agrees that this is the traditional and standard view is Jon Moline in "Aristotle, Eubulides and the
Sorites," Mind, Vol. 78 (1969), pp. 393-405. Moline traces the tradition back to Cicero. For a final example of "light" treatment
consider Friedrich ?berweg's five volume Grundriss der Geschichte der Philosophie (Akademische Drucku. Verlagsanstalt: Graz
1953), which has one-third of a page on the paradoxes I deal with on page 157 of Volume I.
6. William and Martha Kneale, The Development of Logic, (Oxford University Press: London, 1971), pp. 114-15. A further ex?
planation is given by Evert Beth, in Foundations of Mathematics, (Harper & Row: New York, 1966).
7. Theodor Gomperz, in Greek Thinkers?A History of Ancient Philosophy, Vol. II, tr. by G. G. Berry, (John Murray: Lon?
don, 1931), p. 190.
8. Jon Moline, "Aristotle, Eubulides and the Sorites," Mind, Vol. 78 (1969), pp. 393-405.
9. Peter Unger, "There Are No Ordinary Things," Synthese, Vol. 41 (1979).
10. Samuel Wheeler, "Reference and Vagueness," Synthese, Vol. 36 (1975), pp. 367-79, and "On That Which Is Not," Synthese,
Vol. 41 (1979), pp. 155-73.
11. Diogenes Laertius, Lives of the Philosophers, Book II, (Oxford Classical Texts, 1964, p. 101). The translation is from Robert
Caponigri's Gateway Edition, p. 155.
12. Diogenes Laertius, op. cit., O. C. T., p. 102; Caponigri, p. 156.
13. Diogenes Laertius, op. cit., O. C. T., p. 104; Caponigri, p. 157.
14. SeeW. K. C. Guthrie, History of Greek Philosophy, Volume III, (Cambridge University Press: London, 1969), pp. 499-507.
15. Alexander Mourelatos, The Route of Parmenides, (Yale University Press: New Haven, 1970.)
16. Parmenides, Fragment B-5, in Mourelatos, op. cit., p. 230.
17. For instance, Montgomery Furth's "Elements of Eleatic Ontology," Journal of the History of Ideas, Vol. 6 (1968), pp. 111-32,
would fit with what I want to say about the Megarian paradoxes.
18. I. M. Bochenski, A History of Formal Logic, (University of Notre Dame Press: Notre Dame, IN, 1961), p. 131.
19. Alexander Rustow, Der Lugner: Theorie, Geschichte und Aufl?sung, pp. 61-65, (Teubner Verlag: Leipzig, 1910).
20. For example, Saul Kripke, "Outline of a Theory of Truth," Journal of Philosophy, Vol. 72 (1975), pp. 690-716.
21. Saul Kripke, ibid.
22. Alfred Tarski, "The Concept of Truth in Formalized Languages" in Logic, Semantics, Metamathematics, (Oxford University
Press: London, 1956).
23. Alfred Tarski, op. cit., p. 165.
24. Zeno's paradoxes are well summarized in an Introduction to Early Greek Philosophy by John Mansley Robinson, (Houghton
Mifflin Company: Boston, 1968), pp. 127-39.
25. For instance, W. V. Quine, in Word and Object, (The M.I.T. Press: Cambridge, MA, 1960), pp. 183ff.
26. P. F. Strawson, Introduction to Logical Theory, (Methuen & Co.: London, 1960), pp. 173ff.
27. This is John Mansley Robinson's translation. (Op. cit., p. 110.)
28. Steven Boer and William Lycan, "Knowing Who," Philosophical Studies, Vol. 28 (1974), pp. 299-344.
29. Samuel Wheeler, "On That Which Is Not," "Reference and Vagueness," op. cit.
30. Peter Unger, for instance, in "There Are No Ordinary Things", op. cit.